A robust inverse halftoning algorithm based on parameter estimation for AM halftone image

In printing industry, Amplitude Modulation (AM) screening is one of the most important halftone methods to create dot screen, and is used to convert continuous tone image into binary image. On the contrary, in order to get a good visual effect and to do further OCR, we need to remove dot screen by employing inverse halftoning algorithm, however, it is difficult to find a tradeoff between screen suppression and edge details preservation. In this paper, a robust inverse halftoning algorithm based on parameter estimation is proposed, which can deal with the above difficulty perfectly. The proposed algorithm comprises three steps, extraction of screening region, estimation of screening parameters, and inverse transform. Experimental results show that the proposed method is independent with parameters and is robust to obtain good visual performance.